Recenter coordinate from extended simulation

User discussions
1

GROMACS version: 2021.3
GROMACS modification: Yes/No

Dear users,
I performed a 1 ns MD simulation of a protein in water, and used the following commands to increase in 1ns:

gmx convert-tpr -s md.tpr -extend 1000 -o md_2ns.tpr
gmx mdrun -nt 32 -v -deffnm md_2ns -cpi md.cpt -noappend

It creates new files named md_2ns.part0002.*. As my system was in a “dodecahedral” box, I used the following command to recenter the protein:

gmx trjconv -s md_2ns.tpr -f md_2ns.part0002.xtc -o md_center.xtc -center -pbc mol -ur compact

When I tried to look for the RMSD, it showed me the 1ns result, not the 2ns. I’ve compared the 1ns with the 2ns RMSD (I thought it was showing only the extended 1ns), but they were the same

gmx rms -s md_2ns.tpr -f md_2ns.part0002.xtc -o rmsd.xvg -tu ns

Could someone show me where I’m mistaken?

2

You’re only analyzing the second nanosecond. If you want to analyze the full 2-ns simulation time, you need to first use trjcat to write a concatenated trajectory file from md.xtc and md_2ns.part0002.xtc.

3

Thank you @jalemkul
Is it right?
I used:

gmx trjcat -f md.xtc md_2ns.part0002.xtc -o final2ns.xtc

It creates a new xtc file. After, I used:

gmx trjconv -f final2ns.xtc -s md.tpr -o md_center.xtc -pbc mol -center -ur compact

It showed a 2ns RMSD.
To use the trjconv and rms, I included the -s (.tpr file). Should I use the first (md.tpr) or the second (md_2ns.tpr) file?

4

It doesn’t matter. They contain identical information (including the starting coordinates), just a different number of steps for the simulation.

5

ok
Thanks @jalemkul